The National Safety Council (2002) estimates that more than 20 million passenger vehicles in the United States are involved in crashes each year. The exact number of vehicles involved in low-speed property-damage-only crashes is not known because many of these crashes are not reported to police or insurers. Nevertheless, data from U.S. automobile insurers indicate that the over-whelming majority of crashes producing vehicle damage occur at relatively low speeds. Each year more than 8 percent of recent model passenger vehicles have crash damage leading to insurance claims, with an average repair cost per claim of more than $3,000. The median damage amount is about $2,000, and the most common amount is in the $600 to $700 range. Furthermore, about 80 percent of the damage claims have no associated injury claims (
Highway Loss Data Institute, 2003a
,
2003b
). These data show that low-speed crash damage constitutes a large portion of the total costs to U.S. society for repairing crashed passenger vehicles. These costs are huge; in 2003 more than $4 billion was spent to repair 2002 model vehicles alone.
In the United Kingdom, insurers report the total annual cost of motor vehicle insurance is in the region of $10 billion. About 70 percent of these costs are related to damage repairs in low-speed crashes, with an average of $2,000 damage per incident (
Association of British Insurers, 2002
). Reducing vehicle damages in low-speed crashes could have a massive global financial benefit.
Most of these relatively minor crashes involve damage to the fronts and rears of vehicles. Vehicle bumper systems should absorb the energy of these minor collisions, preventing damage to more expensive vehicle body parts. Unfortunately, although bumpers are designed to meet the minimum standards required by the regulatory body for their given market, they still can allow thousands of dollars of preventable damage in minor collisions. Existing test procedures provide no incentives for manufacturers to improve these bumper designs.
There are four main test programs that evaluate vehicle bumper systems worldwide. In the United States, the National Highway Traffic Safety Administration (
NHTSA, 1977
) has established a bumper standard for passenger cars that includes a series of car-into-barrier tests and subsequent pendulum tests on each vehicle. Barrier and pendulum impacts are conducted at 2.5 mi/h (4 km/h) on the full-width of the front and rear bumpers. Pendulum impacts also are conducted at 1.5 mi/h (2.4 km/h) on the corners of the bumpers. Although no damage to other parts of the vehicle is allowed from these impacts, unlimited damage is permitted to the bumper and its attachment hardware. The test pendulum impacts the vehicle at a height between 16 and 20 inches, effectively regulating the bumper heights of passenger cars. There is no NHTSA bumper standard for passenger vans, sport utility vehicles, or pickups. Regulatory tests in Canada are similar but with an impact speed of 5 mi/h (8 km/h) for the front and rear tests and 3 mi/h (4.8 km/h) for the corner impacts (
Transport Canada, 1978
)
In Europe, vehicle manufacturers must comply with the bumper standard specified by the United Nations Economic Commission for Europe (1980). In this test, a profiled pendulum strikes a vehicle at a height of 15.25 to 19.75 inches with a speed of 2.5 mi/h (4 km/h) on the front and rear bumpers. A similar corner test is conducted at 1.6 mi/h (2.5 km/h). The standard does not dictate unacceptable amounts of bumper damage but only ensures the integrity of headlamps, turn signals, and steering controls.
The Insurance Institute for Highway Safety (
IIHS, 2002
) has a consumer-information low-speed testing program designed to evaluate bumper performance based on the cost of repairing vehicle damage resulting from four test configurations. The four configurations are full-width front and rear impacts into a flat barrier, a front impact into an angled barrier, and a rear impact into a centered pole. All four tests are conducted at 5 mi/h (8 km/h). The first two configurations are patterned after the federal full-width impacts, whereas the latter two represent additional crash configurations that may be encountered in the real world.
The Research Council for Automobile Repairs (
RCAR, 1999
) uses a fourth test procedure for assessing bumper designs. It consists of two tests evaluating bumper performance as well as how easily and cheaply any damage can be repaired. The tests are a 40 percent overlap front impact into a flat rigid barrier and a 40 percent overlap rear impact by a 1,000 kg mobile barrier. The test vehicle strikes the barrier at 15 km/h (9.3 mi/h) in the frontal test, and the moving barrier strikes the vehicle at the same speed in the rear test. The RCAR test program is used by several test houses in Europe, Asia, and South America. The Motor Insurance Repair Research Centre (Thatcham) in the United Kingdom uses the results from this test to rate vehicles for insurance pricing purposes.
Although each of these test programs has its strengths and weaknesses, a limitation common to all four is the inability to test for the possibility of bumper override and underride. In real-world low-speed crashes, vehicles frequently underride or override other vehicle bumpers due to height mismatches. Even when bumper heights are relatively well matched and the bumpers initially engage, under/override still can occur due to inability of one or both bumpers to remain in position and stable during the crash.
In 1999, IIHS conducted vehicle-to-vehicle tests that illustrate the bumper stability issue. In these tests, the front of a vehicle moving at 10 mi/h (16 km/h) struck the rear of a stationary vehicle of the same make and model. In the six tests, two models exhibited instability problems: the 1999 Ford Windstar overrode the rear bumper of its stationary partner, whereas the 1999 Saab 9-3 underrode the rear bumper of its partner (
Figures 1
-
2
).
![]()
![]()
Evidence from real-world crashes confirms that bumper instability is a widespread and costly problem. Between November 2001 and February 2002, IIHS surveyed vehicles brought to five drive-in claims centers operated by insurers in the Washington, DC, metropolitan area (IIHS, unpublished data). Professional appraisers estimated the type and amount of damage for 509 vehicles (342 passenger cars) that were in relatively minor low-speed crashes. Approximately one-third of the passenger cars that collided with other vehicles experienced underride or override, and the associated damage was substantially higher for these vehicles. Two-thirds of the cars that struck SUVs sustained underride or override damage, whereas one-fifth of the cars that struck other cars sustained this type of damage. Passenger cars frequently underride SUVs because of height mismatches. However, the types and locations of the damage (such as vertical scrapes on the bumper covers) suggested that when underride or override occurred in car-to-car crashes it usually was after initial engagement of the bumpers.
Thatcham conducted a survey in 1999 (
Doyle, 2000
), in which 1,500 real-world car-to-car impacts were assessed and the frequencies of common body part replacement and repair evaluated. Of 600 vehicles involved in frontal impacts, 86 percent required new bumper covers, 89 percent new headlamps, 42 percent new grilles, and 49 percent new hoods. Only 13 percent required any replacement of structural components such as front cross members, and 5 percent required replacement of frame rails. These findings suggest that during frontal crashes, there is a tendency for the striking car to underride the bumper of the other vehicle, causing more damage to cosmetic parts than to the structural elements typically damaged in low-speed damageability crash tests. Good engagement of structural items would likely cause more structural damage at frame height and less cosmetic damage above frame height.
The instances of override and underride observed in car-to-car tests and the real-world data emphasize the prevalence of the bumper instability problem. Today's low-speed bumper assessment tests provide no incentives to vehicle manufacturers to address this problem. Positive test results against existing barriers and under current test conditions do not guarantee good bumper engagement in the real world. Test procedures are needed to evaluate a bumper's ability to remain in position during a crash. This paper describes initial tests designed to isolate the causes of bumper instability and to establish the requirements of a test procedure for evaluating this instability.
